Last quarter, the enterprise customer we had spent months building toward backed out. Fairly suddenly, right as the paperwork was landing. I had braced for that news to feel like a gut punch. It did, for about twenty minutes.

Then a different problem took over, and it turned out to be the one that mattered. We had a team of around forty people, a good chunk of whom had been building specifically for this customer. We had even lined up a hire against the project. By Monday they would all know, and what we said in the next few hours would shape how they carried it.

Founders obsess over the dead deal and the revenue hole. Both matter. But if you employ people, your highest-leverage move on day one is internal communication, and the most common way to botch it is to treat it as spin. This is the playbook we used when losing an enterprise deal, mapped to the few ideas that actually have research behind them, including the places we went off script.

You are no longer the decision-maker. You are the messenger.

The most useful reframe came early. We did not decide to end this project. The customer did, citing an internal portfolio restructuring, a pivot toward other priorities, and a view that our cost was too high for the scope on offer. They told us fuller feedback was coming in writing.

That put me in an unfamiliar seat: not the founder who usually decides, but the middle manager delivering a decision they did not make and do not fully understand. That is a different job with its own rules. Amy Gallo’s piece on delivering bad news you did not decide is the closest thing to a manual for it.

Why the spreadsheet is the wrong first move

The instinct, and probably yours too, is to run to the financial model. How big is the hole, can we still hit plan, are we still cash-flow positive. Useful questions. Wrong audience.

Your team is not running to the spreadsheet. They are running to one question: am I going to be okay? Until that question is answered, very little else you say gets through. Robert Sutton, whose framework anchors the section below, describes the chronic anxiety that sets in when people cannot predict what happens to them. The deal math is real, but it is an input to your message, not the message itself.

The Reassurance Triad

This is the part we got most right, and the part most worth stealing. Strip the message down and people need to hear three things, in plain language. We came to call it the Reassurance Triad.

Security. The people in the room are safe, and you say so honestly. We stated that no one’s role was at risk because of this, and we were specific about the horizon rather than waving it away. Vague comfort (“don’t worry”) is worse than a dated, honest promise (“nothing changes here for the next few months”).

Stability. The company is fine without the deal, and you show the math instead of asserting it. We walked through the revenue we now expected, the costs against it, and the fact that we stayed operationally in the black. A setback, not a threat to the company’s existence.

Future. There is more coming, and it is concrete. We named the other live enterprise proposals already on the table and the interest sitting behind them. Not a pep talk. Specific evidence that the pipeline did not die with this one.

These three legs are our packaging; the load-bearing ideas are not. They map almost exactly onto Sutton’s four levers in his work on communicating hard decisions: predictability, understanding, control, and compassion. The “future” leg is what Gallo simply calls focusing forward.

One sequencing rule: truth before triad

The triad has a failure mode. If you lead with reassurance, you sound like you are managing people instead of leveling with them. So we opened with the plain fact: the customer had, fairly suddenly, backed out. No cushion, no “but” in the first breath.

Name the bad news plainly. Then the reasons you have. Then reassure. The moment “but here is the good news” arrives too early, trust leaks out. Gallo’s blunt version of this rule is be direct and avoid mixed messages.

How to communicate when you do not have the full story

We did not have the complete why, and we would not for days. The temptation in that gap is to either fill it with speculation or to quietly let the customer carry the blame. Both are traps.

What worked was naming the boundary out loud. Here is what they told us, here are the reasons we have, the full feedback is coming next week, and we will share what we learn. Transparency about uncertainty is itself a form of predictability. A clearly marked gap beats a confidently invented story. It also beats throwing the customer under the bus, which feels good for a second and costs credibility for far longer.

The cascade: who hears it, and in what order

Sequence is not a detail. It is most of the execution, and it rides on the same cross-functional collaboration you hopefully practice on an ordinary week. Ours ran like this:

1. Founders aligned first. My co-founder, our founder’s associate, and I settled on one version of the story before anyone else heard a word. Two versions in the wild is how trust dies.
2. Closest people first, in person. The engineers, designers, and PMs who had built for this customer heard it from me directly, in a room, before it existed in writing. They had earned that.
3. Team leads before their teams. Leads got the written version ahead of the wider company, so no lead was caught flat-footed by their own report knowing first.
4. Whole company, fast, with a channel for questions. We set a near-term all-hands, asked for confidentiality until then, and told leads to come straight to us if they sensed unease.

The principle underneath all four: no one should hear it at the same moment as the people they manage, and the gaps between the circles should be as short as you can make them.

Run the post-mortem, and protect the people who built it

We booked a post-mortem with the people closest to the deal. The framing we put in writing was explicit: this is about learning, not blame, and we need your honest view of what you saw.

That framing is not soft. It is the single highest-leverage thing a leader does for failure. Amy Edmondson’s research on learning from failure found that executives estimate only 2 to 5 percent of failures are truly blameworthy, yet 70 to 90 percent get treated as if they were. That gap is exactly where teams learn to hide things.

Two guardrails. First, de-blame in writing before the post-mortem, not after, so people arrive willing to talk. Second, scope the learning to what you actually controlled: pricing, scope, qualification, how single-threaded the relationship had become. For us, a lot of that traced back to who we should have qualified out sooner, the same trap explored in buyers our market lacks. The customer’s internal pivot was not yours to control, and treating an outside shock as a personal failing is the wrong kind of analysis.

Where we went off the textbook

Models are scaffolding, not scripture, which is the whole point of the framework trap. We leaned on the research above and still got three things less right than the textbook would prescribe. This is the part other founders should learn from most, because we did not run it perfectly.

We left a weekend in the gap. Leads and the closest team knew by Friday. The rest of the company heard it at Monday’s all-hands. With that many people already in the know, the teams who had not been told (sales, customer success, marketing, and operations) were exactly the ones at risk of hearing it secondhand over the weekend. Run it again and we would either shorten the gap to hours rather than days, or hand leads a short holding line for weekend questions.

Control was our weakest leg. We were strong on security, stability, future, and honesty. But “give people a say” showed up only implicitly, in the post-mortem invitation. Giving the team genuine input into what we double down on next would have closed the loop. It is the lever founders skip most often.

We left a contradiction in the message. We repeated the customer’s “your costs were too high” and, separately, told the team they had done excellent work. Both are true. Left unconnected, they invite a quiet “so was it our fault?” The fix is one sentence: the product was strong, the price-and-scope fit for this specific buyer was not, and that is a positioning lesson rather than a quality failure. That sentence also happens to be the honest hook for the post-mortem.

A smaller one worth flagging: the moment you open the books to the whole company, those numbers get scrutinized. Make sure they are airtight before they leave the room, and know that you have set a precedent to keep sharing them.

The first-day playbook

If a deal dies on you tomorrow and you have a team, here is the version you can run before lunch:

1. Align the founders on one story before anyone else hears anything.
2. Write the message in order: bad news plainly, the reasons you have, then the boundary of what you do not yet know.
3. Deliver the Reassurance Triad: security, stability, future, each with an honest specific.
4. De-blame explicitly, in writing.
5. Cascade by circle: closest people in person, leads before teams, all-hands fast, confidentiality plus an open door.
6. Book a blameless post-mortem, scoped to what you controlled.
7. Decide the external move (a compensation ask, leaving the door open) but plan internally as if the deal is closed, so it never becomes a zombie that quietly drains focus. That last discipline comes straight from the HBR argument about why companies cannot kill dying projects.

The deal dying was outside our control. The four hours after it died were entirely inside it. That is usually where the real test of a founder sits. Not in whether you lose the customer, which you sometimes will, but in whether the team comes out of it trusting you more than it did the day before.